#emc | Logs for 2004-10-25

[00:00:39] <jmkasunich> compared to the last couple of weeks, that may be a good thing
[00:00:59] <les> yeah lots of controversy there
[00:01:02] <jmkasunich> I'm sorting thru 400+ emc related emails from the last 2 weeks
[00:01:11] <les> wow
[00:01:40] <les> I had some atlanta hobby metalworking types come up to see emc in action
[00:01:42] <paul_c> bugger me - I don't get anywhere near that many
[00:02:01] <jmkasunich> commit list, board list, dev list, user list, and a few personal
[00:02:12] <jmkasunich> you probably got as many too, just didn't realize it
[00:02:31] <les> I have been getting email ORDERS
[00:02:42] <jmkasunich> mine stay in the webmail mailbox until I fire up a traditional email client and download them, so they are more noticable
[00:02:44] <paul_c> 400 emails for the whole month
[00:02:54] <les> which I could not figure out since I have the mag ads turned off
[00:03:15] <les> then I checked my web page stats
[00:03:24] <les> 7000 hits/month
[00:03:56] <les> so that's it I guess
[00:04:48] <jmkasunich> paul_c: I stand corrected - it's been three weeks - the first ebo mail was on october 1
[00:05:06] <jmkasunich> usually I download them more often
[00:05:13] <les> so ebo has just gotten mad and walked away?
[00:05:20] <jmkasunich> pretty much
[00:05:29] <les> too bad
[00:05:50] <paul_c> Yo Ray
[00:05:55] <les> hey ray
[00:05:57] <jmkasunich> morning ray
[00:06:21] <paul_c> * paul_c has hit problems with the box on Port 32
[00:07:33] <rayh> Morning guys.
[00:07:46] <rayh> What kinds of problems on 32?
[00:07:59] <paul_c> no space left on hda1
[00:08:36] <rayh> That's got to be 40 Gig total.
[00:10:46] <paul_c> hda1 is ~6Gig
[00:11:21] <paul_c> hda2 & hda5 are abot 30Gig a piece
[00:11:28] <paul_c> but not mounted
[00:11:51] <rayh> Want to make a file system on 5
[00:12:21] <les> I ought to run hardware raid in the emc boxes
[00:12:25] <rayh> 2 is the extended
[00:12:39] <les> it's working well on the office box
[00:12:46] <rayh> Hey Les. Didn't see you.
[00:12:53] <les> hey ray
[00:13:57] <rayh> Pretty soon you'll need to clone yourself for all the projects you've got working.
[00:14:23] <les> I am pretty busy
[00:14:30] <les> working 6 days
[00:14:40] <les> need a break
[00:15:27] <rayh> Hey that's what the commandments you're routing say you should do;)
[00:15:38] <les> ha
[00:15:56] <les> well I am trying to hire sanders
[00:16:11] <les> woodworking is at least 75% sanding
[00:16:24] <les> (if you have cnc)
[00:18:25] <rayh> I knew a Sanders once. Long ago in college.
[00:18:50] <les> well I have a job for him haha
[00:19:36] <paul_c> Now that I have nailed a minor bug with the background image....
[00:19:40] <les> I WISH sanding could be cnc automated
[00:19:42] <rayh> Can't make the cnc do the sanding?
[00:19:56] <paul_c> Is there anything else that needs to be added to the CD Ray ?
[00:19:56] <les> but not easy...it's a touchy feely thing
[00:20:28] <rayh> paul_c: Yep. But I'm still changing stuff around for incremental jog.
[00:21:09] <paul_c> OK - Nothing major that can't be added in a bit..
[00:21:11] <rayh> I'm stuck on some array things.
[00:21:48] <SteveStallings> Paul, did the image tweek include my request to be able to see messages through the image during the boot process?
[00:21:54] <rayh> paul_c: How is the revised disk for installing.
[00:22:21] <paul_c> SteveStallings: Different project
[00:22:42] <SteveStallings> Oops, late to the show, wrong story line....
[00:22:50] <rayh> paul_c: It would appear that there is a file system on 5 if you need to use that.
[00:23:01] <paul_c> rayh: Scripted the removal of the Install icon from the Desktop
[00:23:11] <asdfqwega> Heh...I'm probably not going to be popular today...
[00:23:20] <asdfqwega> I've found a problem
[00:23:22] <paul_c> and it launches KDE as the last new user
[00:23:38] <rayh> paul_c: That sounds good.
[00:23:42] <paul_c> EMC starts up as soon as KDE is running
[00:23:56] <rayh> Nice.
[00:24:15] <rayh> asdfqwega: What kind of problem?
[00:24:16] <asdfqwega> M[3/5] G01 Xxxx Yyyyy - works
[00:24:29] <asdfqwega> M[3/5] G00 Xxxx Yyyyy - doesn't
[00:24:37] <paul_c> Stuff like boot graphics will have to wait untill the next kernel/distro generation.
[00:25:07] <rayh> paul_c: We do what we can where we are.
[00:25:24] <les> is there a way to make the emc1 boot graphic full screen?
[00:25:33] <asdfqwega> Putting the M3/M5 on the same line as G00 makes the same problem as the M62 problem I just emailed to the users' list
[00:26:12] <paul_c> les: Use a larger image...
[00:26:37] <les> tried that and it seemed to scale to a small window
[00:27:07] <paul_c> Ray would probably be able to tell you..
[00:27:43] <les> I may have just not had the image large enough
[00:28:08] <rayh> Whatcha doing with the image, les?
[00:28:23] <asdfqwega> I'm going to double check if M62/M63 works properly with G01...
[00:29:06] <les> oh, just a wattsfurniture logo rather than the default Matt Shaver mill picture
[00:29:37] <jmkasunich> the popup is done by a binary called popimage, I dunno if it has a man page or anything
[00:29:37] <les> want to put some stuff on the GUI too if I can muddle through tcl
[00:29:43] <asdfqwega> les & rayh: what about that penguin picture from nekobox.org?
[00:29:48] <rayh> les: You want to use it as a splash.
[00:30:24] <les> I will check on popimage
[00:30:39] <les> penguin picture would be good
[00:30:47] <asdfqwega> [snickers]
[00:30:48] <jmkasunich> line 532 of emc.run is the call to popimage, has the path
[00:31:47] <les> I will check that out...popimage might have some command line parameters or something
[00:31:57] <rayh> I believe that the image is desplayed real size.
[00:32:18] <les> ok I will first try larger images still
[00:32:48] <les> and I will attempt to muck up some of your tcl code haha
[00:33:25] <rayh> If you've got GIMP it has a scaler.
[00:33:27] <paul_c> popimage is a tcl script
[00:33:37] <les> hmm
[00:34:02] <paul_c> 46 lines
[00:34:09] <jmkasunich> didn't know that
[00:34:12] <les> I think GIMP is on Paul's distributions
[00:34:17] <paul_c> in plat/nonrealtime/bin
[00:34:24] <les> k
[00:34:42] <paul_c> convert will also scale images
[00:34:53] <rayh> did not know that.
[00:35:18] <paul_c> if you are a command line junkie
[00:35:40] <les> what about emc waking up full screen?
[00:35:49] <jmkasunich> would display be an appropriate way to do the popup?
[00:35:58] <jmkasunich> instead of popimage?
[00:36:37] <les> Would like "chips" to be on the gui somewhere
[00:37:02] <rayh> asdfqwega: M[3/5] G00 Xxxx Yyyyy works on generic.run here with Live rc46 and a month old EMC.
[00:37:36] <rayh> jmkasunich: display?
[00:37:42] <jmkasunich> man display
[00:37:48] <jmkasunich> may or may not be on your system
[00:37:48] <paul_c> jmkasunich: display would need killing after a delay
[00:38:08] <jmkasunich> was wondering about that - it has a slideshow option, and a -delay option
[00:38:25] <paul_c> but it woud require ImageMagick to be installed
[00:38:36] <jmkasunich> true
[00:39:01] <jmkasunich> maybe that's why Ray doesn't have it - I seem to recall installing IM some time ago
[00:39:28] <jmkasunich> suggestion retracted ;-)
[00:39:47] <paul_c> if Ray has convert, then he'll have display
[00:40:04] <jmkasunich> actually Les is the one who's trying to change his logo
[00:40:21] <rayh> I thought that if you have convert, you have image magic
[00:40:31] <paul_c> yup
[00:40:35] <les> I am more interested in graphics on the GUI
[00:40:41] <jmkasunich> if you have IM, you have convert, display, and a bunch of others
[00:40:50] <les> I like pretty pictures I guess
[00:40:55] <paul_c> Not sure if BDI-2.xx has ImageMagick though
[00:41:13] <les> I am still running that
[00:41:16] <jmkasunich> les - the splash screen is popped up from the run script, hence it's a command line thing
[00:41:29] <les> ok
[00:41:37] <rayh> It's (IM) is on 46
[00:41:45] <les> got it
[00:42:10] <rayh> It's also on knoppix
[00:42:32] <paul_c> BTW John - Cracked a problem that was preventing emc2 compiling cleanly on a 2.6 kernel
[00:42:42] <jmkasunich> that's good news
[00:43:03] <paul_c> However...
[00:43:05] <jmkasunich> is it too early to think about setting up a 2.6 compile farm slot?
[00:43:15] <paul_c> emc2 will not work at present
[00:44:20] <jmkasunich> because?
[00:44:21] <rayh> Um. You'd have to write a script to use display with the startup of emc.
[00:44:36] <jmkasunich> or hack emc.run
[00:45:06] <rayh> And since we've got one... It may be easier to scale the image than to write a new popimage.
[00:45:07] <jmkasunich> not something to do for distribution, but if it's just for your own machine...
[00:45:22] <paul_c> jmkasunich: 2.6 does not export module symbols by default
[00:45:42] <jmkasunich> I see
[00:46:04] <jmkasunich> I wonder if there is gonna be (or already is) a 2.6 aware version of the LDD book?
[00:46:21] <paul_c> there will be, but not yet
[00:46:23] <rayh> les: Send me the stuff you want to put together and I'll make your splash.
[00:46:37] <les> Well if emc ends up on some of my manufacturing stuff in Chicago I might need graphics on the GUI
[00:46:53] <rayh> I can do that as well.
[00:47:11] <rayh> Have you got your logo in digital?
[00:47:14] <paul_c> another thing that you can't do is build two (or more) objects from a single foo.c
[00:47:26] <les> great...thanks..I will get that together
[00:47:40] <les> logos can be in most any picture format
[00:47:51] <jmkasunich> so no more realtime and user space objs from the same sourcefile? that sucks
[00:47:51] <rayh> Matters not to me.
[00:48:13] <paul_c> jmkasunich: No... two or more kernel objects
[00:48:24] <jmkasunich> where do we do that?
[00:48:31] <paul_c> usr & kernel from the same foo.c is still OK
[00:49:05] <SteveStallings> LDD = Linux Device Driver book? If so, keep me posted about availability.
[00:49:19] <jmkasunich> steve, yes about ldd
[00:55:38] <rayh> Hoh, Wah! I've got a prototype jog increment routine that switches inch, mm, degree.
[00:55:54] <rayh> That only took a few days.
[01:00:20] <paul_c> jmkasunich & rayh - With regards the latest posting from Thomas Formella
[01:00:36] <jmkasunich> Thomas who?
[01:00:39] <jmkasunich> * jmkasunich checks email
[01:00:59] <paul_c> We need to organise ourselves to coordinate these things.
[01:01:20] <jmkasunich> the plotter guy?
[01:02:05] <rayh> How so?
[01:02:48] <paul_c> on the board chnnel ?
[01:05:32] <les> wether is clearing up...must go out and deal with leaves
[01:05:41] <les> will stay on and read later
[01:06:20] <les> weather
[01:12:38] <danfalck> * danfalck is back (gone 14:42:49)
[01:12:47] <danfalck> good morning guys
[01:13:19] <rayh> Hi Dan.
[01:13:26] <danfalck> it looks like lots of exciting things are going on
[01:13:43] <danfalck> in the open source world
[01:13:56] <danfalck> You guys are very busy
[01:14:17] <danfalck> Hi Ray
[01:14:19] <rayh> danfalck: What do you see in open source?
[01:14:39] <danfalck> Varkon just put out another release
[01:14:56] <danfalck> probably just a minor thing to a lot of people, but exiting for me
[01:15:17] <danfalck> they mention control of robots in the release notes
[01:15:56] <rayh> Oh. I saw something about it being used as the base code in a couple of rendering or such projects.
[01:16:37] <danfalck> I also noticed here, that the Ebo stuff, controversial as he was got things stirred up
[01:16:48] <danfalck> in a good way
[01:17:19] <danfalck> pain in the ass that he was in the end and all
[01:17:57] <danfalck> I don't think I have ever seen as many emails from the list before...
[01:18:40] <rayh> It was quite a storm.
[01:19:00] <rayh> I think more so off list for some folk.
[01:21:02] <rayh> I heard that someone got more than 300 posts.
[01:21:17] <SteveStallings> Can someone comment on our progress towards automake? I see lots going on including J Stark's commits today.
[01:21:43] <paul_c> automake is not on the agenda at the moment
[01:22:00] <paul_c> a working auto-configure is the main goal
[01:22:22] <SteveStallings> Oops, I meant auto-config
[01:22:23] <rayh> We're speaking of emc2 now?
[01:22:37] <paul_c> yup
[01:23:49] <Imerator_away> Imerator_away is now known as Imperator_
[01:26:26] <jmkasunich> I would really like to keep run-in-place, if at all possible
[01:28:39] <paul_c> That is up for peer review...
[01:29:30] <SteveStallings> Lacking run-in-place, how should developers test code changes?
[01:30:13] <rayh> How about apt-get update.
[01:30:32] <jmkasunich> apt-get is suitable only for debian and related systems
[01:31:52] <jmkasunich> IMO apt-get (like rpm) is on the would be nice list, but can't be the prime method of distribution - source tarballs (users) and CVS checkouts (dev) should be prime
[01:32:03] <rayh> I think that I was suggesting that developers move in that direction.
[01:35:43] <rayh> I really do like run in place and would miss it very badly.
[01:36:06] <rayh> I often get eight to ten of these lined up along side each other in emctest.
[01:36:10] <jmkasunich> I'm not current on the progress of the autoconf branch, and don't understand the technical details
[01:36:34] <jmkasunich> so in this case, I'm the one spouting off about how much I want something, with no idea how difficult it actually is
[01:36:46] <jmkasunich> I need to look into it
[01:37:19] <paul_c> a simple make builds for "run in place"
[01:37:38] <paul_c> make install modifies stuff for a full install
[01:37:53] <paul_c> changing paths where needed
[01:38:03] <jmkasunich> so I thought - but Jon Starks most recent post says run-in-place is now busted
[01:38:55] <SteveStallings> That was why I asked.
[01:39:04] <paul_c> OK... So we drag him in here and get it fixed.
[01:39:51] <jmkasunich> IOW, in your opinion, run-in-place is not "up for peer review", it is a non-negotiable requirement?
[01:40:17] <paul_c> I never said that
[01:40:29] <jmkasunich> "so we drag him in here and get it fixed"
[01:40:38] <paul_c> J.S. said the changes were up for peer review
[01:40:47] <jmkasunich> maybe we're mis-communicating
[01:41:40] <paul_c> I think we both agree that "run in place" is a requirement
[01:41:46] <jmkasunich> I read J.S.'s post as "autoconf would be easier if we allow run-in-place to stay broke, and I want the group to decide how important it is that I/we fix it
[01:41:58] <paul_c> and a "make install" is desireable
[01:42:41] <jmkasunich> I agree
[01:42:47] <jmkasunich> the process should be:
[01:42:48] <jmkasunich> ./configure
[01:42:50] <jmkasunich> make
[01:42:56] <jmkasunich> run in place if desired
[01:42:59] <jmkasunich> make install
[01:43:03] <jmkasunich> run from installed location
[01:43:17] <paul_c> yes
[01:43:30] <jmkasunich> a make that produces a runnable emc in place is not optional
[01:45:09] <paul_c> I feel the default should be "run in place"
[01:45:24] <jmkasunich> I worded my previos lines poorly
[01:45:31] <jmkasunich> trying again:
[01:45:36] <jmkasunich> ./configure
[01:45:37] <jmkasunich> make
[01:45:44] <jmkasunich> run in place for testing
[01:45:50] <jmkasunich> make install
[01:45:56] <jmkasunich> run from installed location
[01:46:15] <jmkasunich> the last two steps are optional
[01:46:43] <jmkasunich> but one goal of autoconf is to make them possible
[01:48:47] <jmkasunich> since J.S. isn't here, should the "peer review" be conducted on list?
[01:49:12] <paul_c> right - But JS may consider instal as the final goal without regard to the other options
[01:49:13] <SteveStallings> Ray, Paul, - update, the web site issues with user admin are fixed.
[01:49:24] <jmkasunich> ie. we post the desired sequence from above (possibly revised) and request comments
[01:50:40] <paul_c> Do we need to post & request comments ?
[01:50:48] <jmkasunich> well post anyway
[01:51:12] <jmkasunich> should "peer review" mean "because John and Paul say so"?
[01:52:29] <jmkasunich> I'd rather it be "because most folks agree that's the best way". or even "because John and Paul convinced most folks to agree its the best way"
[01:53:25] <rayh> SteveStallings: That should make expanding editing easier.
[01:54:30] <rayh> I'm in for ./configure ./make ./make install as three distinct steps.
[01:54:42] <paul_c> At the moment, I see Alex working from one end, J.S. from another, and myself in the middle throwing in suggestions
[01:54:55] <paul_c> and/or changes were needed
[01:55:40] <paul_c> What I don't see is much (if any) communication or coordination
[01:57:03] <jmkasunich> so who should be coordinating it?
[01:57:44] <paul_c> Exactly !
[01:58:19] <jmkasunich> I meant - what person do _you_ think should be coordinating it.
[01:58:32] <jmkasunich> seems like the candidates are you, JS, or Alex
[01:59:35] <jmkasunich> or do you think it's the board's job as a group?
[02:00:19] <paul_c> * paul_c thinks it is certainly a board thing
[02:00:47] <jmkasunich> large scale decisions yes... finer scale not so much
[02:01:04] <jmkasunich> run-in-place is a large scale decision
[02:01:25] <jmkasunich> following a particular file system standard is a large scale decision
[02:01:35] <jmkasunich> where each and every file goes is not
[02:02:38] <jmkasunich> the implementors should be able to figure out most of them from the "follow this standard" decision, and can post if guidance is needed for some files
[02:03:39] <jmkasunich> I also _strongly_ believe in the "document what you plan to do before you do it" school
[02:04:13] <jmkasunich> IMO, the first bunch of commits should be to directory.map, showing where things are expected to end up after both "make" and "make install"
[02:04:27] <jmkasunich> after the dust settles on that, then go ahead and implement it
[02:04:52] <rayh> * rayh believes that the document what you plan is also an evolutionary thing.
[02:06:08] <jmkasunich> several times in the FSSTND thread, I asked that directory.map be updated
[02:08:31] <elson> Hello, Dave!
[02:08:40] <dave-e> hi Jon
[02:09:04] <jmkasunich> Jon, Dave - we've been discussing the implications of Jon Stark's most recent commit on the autoconf branch
[02:09:20] <jmkasunich> (which breaks run-in-place)
[02:09:52] <elson> Run-in-place? Is that running from a CD, or something else?
[02:10:05] <dave-e> thats why it is a branch ;-)
[02:10:12] <paul_c> run from the directory the code was compiled in
[02:10:51] <elson> Ah, that sounds quite useful for testing. Any idea why it breaks?
[02:11:02] <paul_c> Not looked at it
[02:11:26] <jmkasunich> I think JS is focusing on getting make install to work, and perhaps the methods he's using don't support a plain make
[02:11:59] <jmkasunich> my wife just got home from church, time for lunch
[02:12:00] <paul_c> At the moment, I'm more concerned with organisation of these assorted adhoc developments
[02:12:54] <SteveStallings> How can you tell from a commit message that the subject commit applies to a branch other than the main branch?
[02:13:33] <paul_c> look at the version numbers
[02:13:40] <SteveStallings> I had assumed that JS's commit broke something in the main branch.
[02:14:03] <paul_c> non-branch versions will be x.nn
[02:14:15] <paul_c> and branches, x.x.x.nn
[02:16:43] <SteveStallings> So JS's commits were a mixture.
[02:17:09] <paul_c> Don't think so
[02:18:02] <dave-e> so the JS commit can be run from somewhere?
[02:19:04] <paul_c> SteveStallings: All of J.S.'s commits have been on the auto_configure branch
[02:19:08] <SteveStallings> At least one main >>[Emc-commit] emc2/scripts emc.run,1.9,
[02:19:46] <SteveStallings> If I am reading correctly...
[02:19:54] <paul_c> dave-e: It would appear that the JS commit only allows you to run from the --install-prefix tree
[02:20:55] <paul_c> SteveStallings: original version 1.9
[02:21:09] <paul_c> Revised version
[02:21:44] <paul_c> IOW this is the first change to emc.run since the fork point
[02:21:50] <SteveStallings> So this indicates the beginning of the branch? (Please don't ask me to read Red Bean, it is at the office right now.)
[02:22:26] <paul_c> Go get your Red Bean book
[02:22:43] <SteveStallings> 8-)
[02:23:03] <dave-e> or buy another copy ;-)
[02:23:19] <paul_c> In a nutshell, it shows the file on the branch has now changed
[02:24:00] <asdfqwega> Whoo-hoo! It worked!
[02:24:13] <asdfqwega> Where's my scanner...
[02:26:17] <paul_c> Now that emc.run has a branch version, you can monitor the increments to see how fast it mutates
[02:27:29] <paul_c> Hey Jon - Now that you are here...
[02:27:44] <paul_c> is there anything you wanted to discuss ?
[02:28:07] <elson> well, I don't know. I guess your last comment in the linuxcnc.org
[02:28:29] <elson> indicates a new BDI coming out "soon". Any word on thia?
[02:28:51] <elson> any word on that is what I meant to say
[02:29:26] <paul_c> Currently working on an updated Live in conjuction with a third party
[02:29:57] <paul_c> Once it is finished, it will form the basis for the next BDI-Live
[02:32:00] <paul_c> Will probably use the 2.4.27 kernel and rtai-3.1
[02:32:26] <elson> OK, sounds good. Once I got past the hurdle of getting EMC
[02:32:47] <elson> to compile on the BDI-Live! it seems to work very well.
[02:33:29] <elson> I have been fighting the problem of incompatibilities of various PC motherboard parallel ports and IEEE-1284.
[02:33:39] <paul_c> Out of necessatity, I had to make some changes to get it to compile with RTAI
[02:34:02] <elson> I found that Dell computers seem to work best. I've never found
[02:34:05] <paul_c> and these changes were bound to trip a few people up.
[02:34:30] <elson> a Dell desktop that doesn't work (yet).
[02:34:54] <paul_c> Thought IEEE-1284 was supposed to be standard
[02:35:07] <elson> Yes, as for the changes for RTAI, one of the changes made all my
[02:35:58] <elson> PPMC/USC/UPC code not work for about 2 months. it only generated compiler warnings, but would croak when run on a RTAI system.
[02:36:43] <elson> And, of course, until I had a working RTAI system here, I was blissfully unaware that anybody who tried the code from the CVS
[02:37:04] <elson> on a RTAI system would get nowhere - kernel module would not load.
[02:40:39] <paul_c> I can only test for the hardware I have here
[02:41:20] <elson> Oh, of course! (Although this problem was of such a nature you'd actually be able to
[02:41:47] <elson> tell without any hardware. insmod would not load the module due to missing symbols.)
[02:42:44] <paul_c> Hmmm... Add a target to make to load all modules and report on failures...
[02:42:50] <rayh> back to a previous topic. focussing development energy. How about something like the debian manifesto as a talking point. http://ftp.debian.org/debian/doc/debian-manifesto
[02:44:52] <elson> The exact cause was a conditional compile using #ifndef realtimeonly
[02:45:25] <elson> which didn't make the right decision anymore, and included iopl for a real-time component.
[02:46:54] <paul_c> The whole rats nest of #ifdefs needs a (being blunt here) a good kick up the bacside
[02:47:40] <elson> Yes, it is a mess, without much of a pattern to how you use
[02:47:55] <elson> various symbols in different modules.
[02:48:15] <paul_c> Did some work with emcmot. earlier in the week cutting out all the includes
[02:49:21] <paul_c> 10 system headers
[02:49:32] <elson> emcmot.c ? Yes, that's a HUGE file, with gobs of nested ifdefs and includes that may not be needed.
[02:50:18] <paul_c> only 5 essential kernel headers required
[02:50:54] <paul_c> four rtai headers
[02:51:03] <paul_c> and a math.h
[02:51:26] <paul_c> oops.. mat.h & float.h
[02:51:38] <dave-e> ray--having scanned the debian-manifesto it looks like part of it could be shamelessly stolen and only the name changed. ;-)
[02:52:11] <dave-e> actully name(s) changed
[02:54:14] <danfalck> * danfalck is away: going out to the shop...
[02:54:44] <dave-e> I've been away from the shop for 10 days now...and it needs attention. see ya!
[02:55:00] <elson> Not to change the subject, but are there any comments on threading?
[02:55:56] <paul_c> my work has been on hold waiting for an encoder/motor to turn up
[02:56:16] <asdfqwega> Man...how did that happen?...wish I had a page counter.
[02:56:57] <elson> Paul: What do you mean? I could send you one of these little servo motors I am using on a Taig retrofit.
[02:56:58] <asdfqwega> I was going to show off a scan of some laser etching I just did with EMC, but my host says I've exceed my transfer limit.
[02:57:37] <elson> Arrgh - get an ISP that doesn't have those pesky limits!
[02:58:01] <les> what is your transfer limit?
[02:58:09] <asdfqwega> 50MB a day
[02:58:32] <asdfqwega> and my nick's not Arrgh - that's a castle ;)
[02:58:58] <les> oh that is small I think I get 30 gig/mo
[02:59:05] <paul_c> * paul_c signed up for an "unlimited service"
[02:59:26] <les> cost a lot?
[02:59:38] <paul_c> �23/month
[02:59:43] <paul_c> �25/month
[02:59:54] <asdfqwega> some search engine must have picked up my little page
[03:00:03] <les> the 30G is $9
[03:00:40] <les> actually that is a mistake...just checked and it's 40G
[03:00:43] <asdfqwega> That'd be nice...except then I'd feel like I'd have to actually use it :P
[03:01:27] <les> well it's a commercial site...but cheap is cheap whether you need that much or not
[03:01:44] <les> now I do not need the 500 email addresses
[03:02:05] <les> I actually gave my cat an email address.
[03:02:17] <les> congo@lmwatts.com
[03:02:21] <les> haw
[03:02:41] <paul_c> how much spam does the moggie get ?
[03:03:10] <asdfqwega> heh...start sending you cat spam..."Hot Dog Action", or somesuch
[03:03:20] <SteveStallings> Suspect moggie would prefer tuna. 8-)
[03:03:42] <les> I was going to give the chickens email too...but they don't have names
[03:04:44] <les> could give them chat rooms...They could peck to each other on the keyboard
[03:04:51] <asdfqwega> Heh...my account's fine, it seems I uploaded the picture to the wrong place.
[03:05:02] <les> oh good
[03:06:00] <rayh> If you sent a copy to Steve we could hammer his server rather than yours.
[03:06:38] <SteveStallings> Gee thanks Ray...
[03:06:40] <asdfqwega> ...
[03:06:46] <asdfqwega> Well, that makes not sense
[03:07:01] <asdfqwega> I can pull up my page, but not the picture.
[03:07:56] <asdfqwega> They probably have some kind of filter to prevent bandwidth leeching
[03:07:58] <les> I thought about a blog for showing pictures and stuff
[03:09:05] <les> that is about all it would be good for at least for me
[03:16:59] <asdfqwega> http://www.bright.net/~janfritz/CNC/An%20example%20of%20laser%20etching%20with%20EMC.html
[03:17:46] <paul_c> asdfqwega: You used M$ to upload didn't you
[03:18:23] <asdfqwega> paul_c: Them's fightin' words, mister.
[03:18:50] <asdfqwega> No, I just rattled off a quick page with Mozilla composer to put the .jpg link into.
[03:19:39] <paul_c> way too many %%20%%20 foo
[03:19:40] <asdfqwega> I have as little contact with M$ stuff as possible.
[03:20:04] <asdfqwega> Yeah, I forgot about how the page title become the file name.
[03:20:18] <asdfqwega> Does it work?
[03:21:02] <paul_c> can you do me a batch ?
[03:21:28] <asdfqwega> ?
[03:21:50] <asdfqwega> How many is a batch?
[03:22:01] <paul_c> more than one
[03:22:06] <asdfqwega> :P
[03:22:29] <paul_c> 10^6
[03:22:44] <asdfqwega> [falls over]
[03:23:28] <asdfqwega> This was a test image I ran through my img2laser script
[03:23:43] <asdfqwega> I did the run on a piece of cardboard
[03:23:56] <paul_c> One would be nice to send to 'lil Sis
[03:24:59] <asdfqwega> I could do them up on wood...which is what I was going to do next, to see how the quality would turn out.
[03:28:18] <asdfqwega> It's 4"x3.5" or about 100mm x 87mm
[03:28:19] <jmkasunich> I knew a guy some years ago who did similar things with a laser on wood
[03:28:56] <jmkasunich> what I remember most distinctly about them was the smell - even days afterwards, they had an unpleasnt charred wood smell
[03:29:19] <jmkasunich> I suppose a lot depends on the type of wood
[03:30:11] <rayh> Heard about a fellow near me (youpper) that cut basswood for model aircraft parts.
[03:30:32] <rayh> The resin turned dark along the path.
[03:30:45] <jmkasunich> that's the word I was looking for - burned wood resin
[03:30:48] <asdfqwega> paul_c: seriously, how many would you want?
[03:31:20] <asdfqwega> paul_c: and what would you like it on?
[03:33:17] <asdfqwega> * asdfqwega prods paul
[03:33:23] <jmkasunich> ok
[03:33:25] <jmkasunich> ops
[03:34:12] <elson> I've seen this at fairs and such with a raster-scanning process, but presumably the way you do it there are no raster lines in the finished product.
[03:35:28] <asdfqwega> That's the theory, but it ain't so...my example is 0.010" resolution, so if you look close you can see the jaggies
[03:35:44] <rayh> ads: nice job.
[03:35:50] <rayh> oops
[03:35:59] <rayh> asdfqwega: Nice job.
[03:36:05] <asdfqwega> thanks
[03:36:30] <rayh> How much power will the laser put out?
[03:37:08] <les> I made some big raster scan machines many years ago
[03:37:11] <asdfqwega> Mine is presumably 15 Watts...and to do the example, I turn it down to half power
[03:37:36] <rayh> Okay. The one near hear cutting basswood was 25 max.
[03:38:10] <les> I had some problems with getting rid of the wood vapors
[03:38:11] <asdfqwega> If you have the right optics, you get a nice focal range for cutting.
[03:38:28] <les> finally had to make an afterburner
[03:39:36] <rayh> Catalytic converter?
[03:40:01] <asdfqwega> les: When I'm on full power, wood smoke becomes noticable after a while
[03:40:07] <les> no...a gas pilot light thing
[03:40:26] <asdfqwega> Me, I'd just open the doors :P
[03:40:41] <les> this was 2000 W so the creosote buildup was unmanageable
[03:41:21] <asdfqwega> What was your feed rate?
[03:41:38] <A-L-P-H-A> hello... anyone ever tinker with PLC or PICs? I'm thinking of doing something like the gecko2000/2002 (I forget)
[03:41:54] <les> gosh...let me remember...about 2 meters/s
[03:42:07] <asdfqwega> Yikes
[03:42:10] <A-L-P-H-A> it's essentially, a pulse generator for you, that deals with the intricacies of controlling multiple axises...
[03:42:29] <paul_c> A-L-P-H-A: There is a PIC assembler/simulator in the Debian archive
[03:42:55] <les> sharply focused for clean carving, defocused some for black charcoal appearance
[03:42:56] <A-L-P-H-A> I'm really not familiar with PIC, I have though, dabbled with atmel controllers.
[03:43:00] <asdfqwega> I'm lucky to operate at .2 m/s
[03:43:08] <rayh> A-L-P-H-A: Matt Shaver had been keeping up with that 2000 project.
[03:43:50] <A-L-P-H-A> I'm subscribed to the geckodrive yahoo list, since I think message 5? :) But I haven't been keeping up. Been busy with other stuff (you guys understand)
[03:44:24] <asdfqwega> * asdfqwega needs to eat, BBL
[03:47:51] <paul_c> A-L-P-H-A: Some of the Atmel processors (arm core) will run embedded linux
[03:52:40] <A-L-P-H-A> paul_c, thanks.
[03:55:29] <paul_c> Just don't ask me how to set up a cross-development box
[03:57:03] <SteveStallings> Time for some yard work. Later....
[04:01:34] <paul_c> * paul_c disappears for Tea.
[04:07:33] <A-L-P-H-A> oh.. TEA... that is such a good idea.
[04:12:21] <alex_joni> hello
[04:12:57] <rayh> Hi alex.
[04:13:11] <alex_joni> hi ray
[04:13:13] <rayh> You made it to the customers place?
[04:13:30] <alex_joni> I'll be there first thing next morning
[04:13:40] <alex_joni> but I made it to his city..
[04:14:51] <alex_joni> what did I miss?
[04:15:11] <rayh> Everything!
[04:15:27] <alex_joni> * alex_joni is confused
[04:15:44] <alex_joni> thank god for my logger .. I'm catching up in a moment
[04:16:19] <rayh> check out asd's link.
[04:19:19] <alex_joni> what link?
[04:19:44] <rayh> <asdfqwega> http://www.bright.net/~janfritz/CNC/An%20example%20of%20laser%20etching%20with%20EMC.html
[04:20:03] <rayh> He made that this morning with his laser.
[04:20:48] <alex_joni> I didn't get there... boy you sure talked a LOT
[04:20:49] <alex_joni> ;)
[04:20:50] <picnet> hi ray, cats dont fart as women dont, they just let out a silent one and cough.
[04:21:04] <alex_joni> hihi
[04:22:01] <picnet> im getting no where fast with my steel gantry construction - lack of tools lack of space and no time to make the parts, im thinking of switching to aluminium profile (again) to get the job finished.
[04:22:20] <picnet> mechaano.
[04:22:48] <rayh> aluminium is easier to work with as long as you can keep the temp stable.
[04:23:11] <picnet> 23c constantly in the room.
[04:23:30] <picnet> plus or minus 5c :)
[04:24:05] <rayh> Around these parts we go from -40 to +100 F.
[04:24:14] <picnet> Indoors?
[04:24:48] <Imperator_> A-L-P-H-A: about your project with the pic: in a german Forum they have made something like a gecko with a ATMEL
[04:25:12] <rayh> Out. With the price of heating fuel it may get a bigger range indoors.
[04:25:58] <picnet> fuels cheap in the us.
[04:26:25] <picnet> http://www.neilturner.me.uk/2004/May/19/fuel_prices.html
[04:27:12] <alex_joni> Imperator: could you send me a link?
[04:29:15] <Imperator_> http://5128.rapidforum.com/topic=112282918267
[04:30:13] <Imperator_> but i think it is not realy a servo controller, it's more a motor controller. But they are all happy with it
[04:30:33] <Imperator_> some have problems to solder it together
[04:30:43] <alex_joni> strange...
[04:30:52] <alex_joni> still reading through my log... man you talked a LOT
[04:31:39] <alex_joni> A-L-P-H-A: stay away from PIC's (this is my pure personal advice)
[04:31:56] <alex_joni> go anything else (C51, HC11, HC12, ATMega, etc.)
[04:32:02] <picnet> German cnc forum?
[04:32:25] <Imperator_> jep
[04:32:35] <picnet> what do you think of the EASGMBH machines ?
[04:32:47] <picnet> http://www.easgmbh.de/
[04:32:51] <Imperator_> aluminium hell
[04:32:55] <picnet> lol
[04:33:06] <alex_joni> jmk, paul_c: what got decided regarding autoconf, make install, paths, directory.map?
[04:33:13] <picnet> novel way of creating a linear using the bosch rexroth stuff.
[04:33:42] <picnet> do you know of any companies in .de land that make gantry's (cheap)
[04:33:58] <Imperator_> we are also searching
[04:34:29] <Imperator_> because we want to build a gantry with 1700x700x300mm range
[04:34:54] <picnet> why the 300mm Z ?
[04:35:10] <picnet> do you have 300mm tools?
[04:35:38] <Imperator_> we think thats enough
[04:36:11] <picnet> mines 1000 x 700 x 100, thats built to the room size, any bigger and I wont get in the room :-)
[04:36:33] <Imperator_> :-)
[04:37:24] <alex_joni> did I tell you guys I ran a M$ TkEMC with emc2 last night?
[04:40:27] <Imperator_> picnet: thats our X-Axis http://cgi.ebay.de/ws/eBayISAPI.dll?ViewItem&category=65448&item=3846861644&rd=1&ssPageName=WDVW
[04:41:17] <alex_joni> guess things get discussed on the board-channel
[04:41:55] <alex_joni> here it's pretty quiet right now...
[04:41:58] <picnet> Thanks for the link I just bid 500 euro.
[04:42:13] <picnet> * picnet <snickers>
[04:42:16] <Imperator_> that seller had three
[04:42:39] <Imperator_> the first was gone for 690 EUR but it was 200mm shorter
[04:42:43] <picnet> 40kg
[04:42:55] <Imperator_> ours was the same than this and was gone for 400 EUR
[04:43:42] <Imperator_> you can download the manual at bosch Rexroth
[04:44:16] <picnet> how many do you have so far?
[04:44:59] <Imperator_> we need one
[04:45:12] <robin_sz> 'ello girls ...
[04:45:13] <alex_joni> hey robin
[04:45:18] <alex_joni> hello zwisk
[04:45:21] <zwisk> greetings!
[04:45:22] <Imperator_> hi robin
[04:45:23] <picnet> nice I hope you win it, looks an amazing bit of kit - >4K euro new?
[04:45:27] <robin_sz> * robin_sz waves
[04:45:56] <picnet> * picnet pretends not to be here (because lets face Im never here)
[04:45:58] <alex_joni> zwisk: you did make some waves with your latest commit on auto_conf
[04:46:08] <zwisk> Did anybody mind that I broke run-in-place in the emc checkin?
[04:46:14] <zwisk> ah..yes... I was afraid I might.
[04:46:29] <zwisk> I figured I'd check it all in in a big batch, so we could easily back out of it if need be.
[04:46:31] <robin_sz> * robin_sz ponders dleving into CVS ...
[04:46:35] <alex_joni> well.. if you care about what was discussed,I could give you a link
[04:46:40] <zwisk> There's some discussion that needs to be had around it all.
[04:46:47] <zwisk> Yes please...
[04:46:51] <Imperator_> Picnet: http://www2.rexroth-star.com/INTERNET/WEBDOWNLOAD.NSF/c48d4357490aec2841256a0d0047028f/33b3c11a1669b89241256b870052f074/$FILE/82571_0699_SFT.pdf
[04:46:54] <zwisk> thanks...
[04:46:55] <alex_joni> logger_aj, bookmark
[04:46:55] <alex_joni> See
[04:47:12] <Imperator_> picnet: http://www2.rexroth-star.com/INTERNET/WEBDOWNLOAD.NSF/4cacbb0890414ccf41256a0200531057/353650bd57f24f6b41256cd3005898fc/$FILE/RD82501_0400_TKK.pdf
[04:47:46] <alex_joni> zwisk: got it?
[04:48:40] <zwisk> yep. thanks... catchin' up at the moment...
[04:50:08] <picnet> imp: thanks just looking, very nice.
[04:51:20] <zwisk> ok... so, I'm not sure anyone figure the chanes I checked in out..
[04:51:37] <zwisk> Basically, I really like the idea of run-in place, but the problem si that the cvs structure doesn't match the install structure.
[04:51:52] <zwisk> So, I have a 'make run' target that creates at 'test' directory and installs into the test directory for running...
[04:52:10] <zwisk> It's not a 'real' install, and allows you to do simply 'make run' out of the box and be up and running...
[04:52:16] <zwisk> but it isn't 'run in place', which I sort of like.
[04:52:21] <alex_joni> short question:
[04:52:25] <zwisk> The challenge is to figure out if there is a way to do both.
[04:52:26] <alex_joni> where do files get compiled?
[04:52:38] <alex_joni> or copied?
[04:52:41] <zwisk> files get compiled the same way (and in the same place) they were copiled before.
[04:52:54] <alex_joni> well.. then you can run them from there
[04:52:57] <alex_joni> after make
[04:53:03] <zwisk> After compiling, just as they did before, make 'test' creates a test directory and populates it with all the files.
[04:53:15] <alex_joni> and with make install copy them to final install point
[04:53:23] <robin_sz> * robin_sz falls of his chir
[04:53:27] <robin_sz> chair
[04:53:33] <zwisk> No, you can't run it from there, because it's so heavily script based. The scripts have relative and not realtive links that need fix up.
[04:53:33] <alex_joni> yes... but after compiling...
[04:53:39] <alex_joni> I see
[04:53:50] <zwisk> yes, make install copies them into the filesystem.
[04:53:59] <picnet> been on the absynth again robin?
[04:54:08] <alex_joni> picnet: lol
[04:54:29] <robin_sz> no .. just suprised someone suggested make test && make install ... will wonders never cease?
[04:54:35] <zwisk> The main advantage to running in place, to me, is the ability to quickly edit scripts and then run them.
[04:54:42] <Imperator_> picnet: we where surprised if we have removed the bellows (is that the englich word ) , because the rails are acurrency SP that is very good and very evpensive
[04:54:59] <zwisk> I'm not sure how to keep this, however, given that the scripts often include things from stuff like 'tcl/scriptname' and the install directory we spoke of has no 'tcl' in it...
[04:55:03] <picnet> robin, all as root of course.
[04:55:47] <robin_sz> well ... I was sort of hoping .. naaah ...
[04:56:04] <robin_sz> just make test would be nice though ...
[04:56:49] <alex_joni> zwisk: I like run-in-place because I can test emc before installing it
[04:56:54] <alex_joni> to see if it works, etc
[04:57:00] <alex_joni> after that I can install it...
[04:57:10] <zwisk> you can still do that. You can type 'make run' right out of cvs and it should compile and run.
[04:57:35] <alex_joni> that's what people where talking about
[04:57:43] <alex_joni> cvs up
[04:57:47] <alex_joni> ./configure
[04:57:48] <alex_joni> make
[04:57:50] <alex_joni> make run ?
[04:57:55] <alex_joni> and run emc
[04:58:02] <alex_joni> after a while... maybe
[04:58:06] <alex_joni> make install
[04:58:07] <zwisk> THe down side, though, is if you edit a tcl script in the sources, you need to remake the 'test' directory.
[04:58:42] <zwisk> That's all good. (I actually ocassionally wonder if just 'make' shouldn't run configure if it needs to as well...)
[05:00:56] <zwisk> The scripting path stuff definitely needs some cleanup still. There's a lot of extra variables being set for a bunch of directories all over the place that could be consolidated once we're sure we like the install.map locations...
[05:02:07] <zwisk> I'd like to create an 'make rpm' target as well. Should that go in this branch? Anybody have preferences for where in the cvs tree package specific things like rpm spec files, debian files, etc should live?
[05:02:52] <alex_joni> I think that needs to be discussed with the board
[05:03:01] <alex_joni> or at least jmk, paul_c, etc.
[05:03:16] <zwisk> cool...
[05:04:57] <alex_joni> but they seem to be busy on the emc-board channel ;)
[05:05:42] <rayh> I'm here. lotta good I do in this discussion.
[05:06:14] <rayh> My biggest concern is the ability to build and test one without breaking an existing one.
[05:06:22] <alex_joni> exactly
[05:06:37] <Imperator_> picnet: you have it,or ???
[05:06:39] <alex_joni> * alex_joni is digging for tcl82.lib
[05:06:56] <picnet> sorry imperator do I have what?
[05:07:35] <Imperator_> the linear guide
[05:08:26] <alex_joni> anyone who would know how to create a .lib from a .dll ?
[05:08:29] <alex_joni> no?
[05:08:33] <alex_joni> thought so...
[05:08:36] <robin_sz> yes ...
[05:08:46] <zwisk> another interesting topic for discussion would be the use of an automated documentation system, such as doxygen. But I'll keep my earth shattering suggestions to only a few at a time for now :)
[05:08:46] <alex_joni> yes?
[05:08:48] <robin_sz> first delete your doze partition
[05:08:53] <zwisk> haha..
[05:08:56] <alex_joni> funny...
[05:09:01] <robin_sz> and shout "dlls get thee away .."
[05:09:04] <alex_joni> I'm actually trying to compile emc
[05:09:11] <robin_sz> on doze?
[05:09:11] <alex_joni> if you can believe that
[05:09:14] <alex_joni> yup
[05:09:18] <robin_sz> prevert
[05:09:19] <alex_joni> rcslib is there ;)
[05:09:35] <picnet> imp: yes saw the linear, no i dont own one.
[05:09:41] <alex_joni> at least emcsh,
[05:10:11] <robin_sz> now .. that I can see as almost useful
[05:10:12] <alex_joni> robin: I ran last night TkEMC from doze wireless to a emc2
[05:10:14] <Imperator_> picnet: because a guy from sweden has bought it.
[05:10:25] <alex_joni> ;)
[05:10:37] <alex_joni> I was really impressed
[05:10:37] <picnet> Im in Finland, dont mention the swedes.
[05:10:49] <Imperator_> ah, sorry
[05:11:02] <alex_joni> the doze stuff was pretty old (a zip from 2002 or so)
[05:11:12] <alex_joni> but still worked on emc2 with autoconf ;)
[05:12:20] <picnet> so holte ein schwedischer brauner Bohrung Ritter es?
[05:12:38] <rayh> emcsh is designed to compile under MS so that folk can run tkemc there.
[05:13:04] <alex_joni> yup
[05:13:10] <Imperator_> :-)
[05:13:42] <picnet> you didnt big higher?
[05:13:46] <picnet> bid even.
[05:13:48] <alex_joni> how about mmio, motsim, mmtask ?
[05:14:05] <picnet> tea & toast time, bbl.
[05:14:09] <Imperator_> picnet: we have already bought one from that guy
[05:14:19] <rayh> I saw an EMC that Bert Eding in NL compiled for NT several years ago.
[05:17:04] <alex_joni> realtime? or just sim?
[05:17:16] <rayh> Realtime.
[05:17:47] <Imperator_> about EMC and Windows: I think it is better to port the Windows guys to Linux :-)
[05:17:55] <rayh> Bert just posted a note to the user list you could ask from that.
[05:18:36] <rayh> I agree. Don't run MS here except for wife's classroom.
[05:18:44] <alex_joni> hmmm.. gotta browse the user list ;)
[05:19:22] <Imperator_> Windows is ok, i work also with it, but writing software makes no fun with Windows
[05:20:23] <alex_joni> hihi
[05:20:31] <alex_joni> just point and click and curse :))
[05:21:01] <rayh> I know that feeling. Used to wear a gold cross and place it in front of the box.
[05:21:15] <Imperator_> for windows everythink is so expensive, also the documentation
[05:21:40] <picnet> depends which edonkey client you run.
[05:21:45] <Imperator_> i have used the kernel hacking book for writing my driver, that is online
[05:21:56] <Imperator_> :-)
[05:22:07] <rayh> * rayh Gotta get outa here.
[05:22:25] <Imperator_> i want to switch to the legal way, sometimes
[05:22:38] <Imperator_> step by step
[05:23:45] <alex_joni> well.. it's pretty pointless
[05:24:44] <jmkasunich> just reading back (sorry, was on the board list)
[05:25:01] <jmkasunich> we would much prefer:
[05:25:04] <jmkasunich> ./configure
[05:25:05] <jmkasunich> make
[05:25:22] <jmkasunich> (result is runable - all paths point to compile directories)
[05:25:25] <jmkasunich> make install
[05:25:39] <zwisk> I agree that this is desireable.
[05:25:40] <jmkasunich> (result is installed - paths are changed to install directories
[05:25:42] <zwisk> Any idea how to do it'?
[05:26:11] <zwisk> One way would be to make the cvs tree more closely resemble the install tree.
[05:26:19] <jmkasunich> not in detail - I've been so busy with the compile farm that I haven't even looked at the autoconf stuff on my main box yet
[05:26:33] <zwisk> (The compile farm is very cool, by the way.)
[05:26:43] <alex_joni> * alex_joni seconds that
[05:26:56] <zwisk> How would you feel about having the compile farm do nightly builds for download (tars, deb, and rpm) as well one day?
[05:27:06] <jmkasunich> * jmkasunich repeats his earlier request to update directory.map, both with desired CVS tree changes if any, and with the proposed install map
[05:27:20] <jmkasunich> don't see why not
[05:27:21] <zwisk> install.map has been added to the tree ....
[05:28:02] <jmkasunich> when? I don't recall seeing the commit msg... sorry about that ;-/
[05:28:25] <zwisk> last night, as part of my huge package of stuff that I wanted all together, in case we decided to back out of it and throw it away :)
[05:29:18] <jmkasunich> ok, I need to take a closer look at that
[05:29:27] <jmkasunich> in the throes of something else right now tho
[05:29:45] <jmkasunich> (several something elses actually - I hate nested interrupts when the happen to me)
[05:29:53] <zwisk> understood! :)
[05:30:08] <alex_joni> * alex_joni suggests asm { cli }
[05:30:30] <jmkasunich> life doesn't work that way
[05:30:40] <jmkasunich> levels of nesting:
[05:30:45] <robin_sz> arrghh! .. he said that word again
[05:30:53] <jmkasunich> compile farm FTP problem
[05:31:02] <picnet> compile farm? - what machines are you running for that?
[05:31:09] <jmkasunich> board stuff, related to setting up trackers for bugs and features
[05:31:12] <alex_joni> yup.. that's a tricky one :(
[05:31:13] <jmkasunich> and this discussion
[05:31:44] <robin_sz> * robin_sz suggests bugzilla or 'RT' for bugs ...
[05:31:50] <alex_joni> I actually liked the way emc did things...
[05:32:02] <jmkasunich> picnet: see http://linuxcnc.org/compile_farm/
[05:32:02] <robin_sz> RT is my prefference for bug/feature tracking
[05:32:05] <alex_joni> I'm not very fond of having a lot of files in /usr/local/bin
[05:32:25] <jmkasunich> too bad robin - we're going with the SF trackers ;-)
[05:32:57] <robin_sz> http://www.bestpractical.com/rt/
[05:33:10] <jmkasunich> the hardware is an _old_ blade system, 8x200MHz Pentiums, 128M ram and 2-4G disk each
[05:34:14] <picnet> one day I'll actually get around to building emc
[05:34:53] <robin_sz> I think the SF trackers are based on a early version of RT anyway ... one of the early python variants from what i recall someone saying ...
[05:35:40] <robin_sz> still, you should at least look at RT before making your minds up for good. admittedly sticking with SF has its advantages though
[05:36:24] <jmkasunich> the project is hosted at SF - there would have to be an overwhelming reason to go outside SF for any project related items
[05:36:43] <robin_sz> sure, you just have to weigh up thte
[05:36:49] <robin_sz> the pros and cons ...
[05:37:10] <jmkasunich> we're having a hard enough time managing the number of choices we already have, at some point you simply have to stop considering other options
[05:37:27] <robin_sz> whatever, I use RT for stuff because its best of breed, SF may well be perfectly good enough
[05:37:37] <robin_sz> and not re-inventing wheels is always a good thing
[05:37:50] <jmkasunich> we don
[05:38:06] <alex_joni> SF can handle the job...
[05:38:20] <alex_joni> it's better to have all in one place
[05:38:22] <jmkasunich> we don't even use 10% of the capabilities of SF - we don't need more capability, we need to use what we have more effectively
[05:38:25] <alex_joni> cvs, lists, trackers
[05:38:26] <alex_joni> etc
[05:38:46] <alex_joni> * alex_joni : btw, I got to compile tcl82.lib ;)
[05:39:06] <alex_joni> now I need tk82.lib :(
[05:39:14] <robin_sz> fair point
[05:39:35] <robin_sz> right .. whose got a spare nitorgen generator to give me?
[05:39:55] <robin_sz> and a 300psi pump please?
[05:40:18] <alex_joni> where do you want to go today?
[05:41:38] <alex_joni> zwisk: let's discuss things
[05:42:11] <zwisk> k...
[05:42:14] <zwisk> I'
[05:42:32] <zwisk> m looking at a stange build error on the cvs stuff I rechecked out...
[05:42:50] <alex_joni> make?
[05:43:11] <zwisk> seems CVS directories are not getting ignored and are getting in the way when things get copied for the intall stuff.
[05:43:18] <zwisk> Silly me.
[05:43:23] <zwisk> Anyway, what shall we discuss?
[05:43:32] <alex_joni> how do you see things...
[05:43:38] <alex_joni> what can, should be done
[05:43:39] <jmkasunich> * jmkasunich listens
[05:44:08] <alex_joni> let's say the board figures out a directory.map
[05:44:11] <zwisk> I like the idea of being able to run out of the source directories without copying. I like this idea a lot. It makes things simpler.
[05:44:20] <alex_joni> yup.. same here
[05:44:29] <alex_joni> what needs to be fixed to be able to do that?
[05:44:35] <jmkasunich> the board agrees (we were discussing it offlist)
[05:44:50] <jmkasunich> two ways to look at the problem
[05:45:00] <zwisk> script paths need to be fixed. There are a bunch of ways to do this, but which one to do is the big question requiring some debate / thought.
[05:45:39] <jmkasunich> 1) make first - make should make something that runs in place, then make install is a secondary project, and it changes things as it is copying
[05:45:57] <jmkasunich> 2) make install first - figure out where everything is gonna go, then make the CVS look like that
[05:46:19] <jmkasunich> Paul and I were doing (1), to the extent that we hadn't even started make install yet
[05:46:28] <jmkasunich> seems like you guys are doing (2)
[05:46:34] <zwisk> Some conventions could really help us out.
[05:46:49] <alex_joni> how about a make = make install --bindir=$EMCDIR/bin --rtdir=$EMCDIR/rtlib etc
[05:47:10] <jmkasunich> huh?
[05:47:16] <jmkasunich> the current
[05:47:18] <jmkasunich> oops
[05:47:25] <zwisk> One problem si that the EMC2_HOME_DIR path is recreated in many scripts. If we move that and all path specific stuff to a single config file, life gets easier.
[05:47:37] <alex_joni> yup... that's a MUST
[05:48:07] <alex_joni> maybe have two sections in that file
[05:48:13] <alex_joni> one for run-in-place
[05:48:16] <alex_joni> and one for install
[05:48:23] <jmkasunich> recreated as in "deduced by looking around"?
[05:48:44] <zwisk> if it's all in one place, sed, autoconf, and other tools exist to make the changes required for installed vs run-in-place paths.
[05:48:52] <jmkasunich> here's a thought
[05:49:08] <jmkasunich> the neccessary paths are in rtapi.conf, hal.conf, and emc.conf
[05:49:16] <alex_joni> yup
[05:49:22] <jmkasunich> the initial copies of those files are in {CVSDIR}/etc
[05:49:30] <zwisk> Yes, deduced by looking around... or using the script path that was used to call the script to create relative paths. Or by picking up one, like EMC2_DIR, and assuming EMC2_BIN_DIR = EMC2DIR/bin ...
[05:49:34] <jmkasunich> they get modified during the install and copied to /etc
[05:49:52] <alex_joni> maybe a rtapi.conf.in
[05:50:03] <zwisk> I created rtapi.conf.in hal.conf.in and emc.conf.in for exactly that reason.
[05:50:07] <zwisk> THey are in cvs now.
[05:50:07] <jmkasunich> ok
[05:50:09] <alex_joni> seen that
[05:50:11] <alex_joni> ;)
[05:50:11] <jmkasunich> so far so good
[05:50:29] <zwisk> There is still a lot of cleanup to be done for that, though.
[05:50:34] <jmkasunich> if it uses the ones in {CVSDIR}/etc, it uses the freshly compiled directories
[05:50:46] <jmkasunich> if it uses the ones in /etc, it uses the installed ones
[05:51:06] <zwisk> right...
[05:51:28] <jmkasunich> now just have to revise the scripts so they check for <current_working_dir>/etc first, then for /etc
[05:51:32] <zwisk> SO, should things like grep, sed, ps, kill, whoami, awk, insmod (there's more but I'll stop there) be searched for at runtime, or found at configure time and added to the config file?
[05:51:54] <jmkasunich> if you run from {CVSDIR}, the first check passes and you use the local ones, else it fails, and you use the installed ones
[05:52:29] <jmkasunich> those paths won't change during install - search at config time I think
[05:52:57] <zwisk> ok. That sounds good to me. It makes many of the scripts much shorter and simpler as well...
[05:53:01] <jmkasunich> for those, run-in-place and run-installed are the same
[05:53:06] <zwisk> yes.
[05:53:11] <jmkasunich> the ones that change are the stuff you are actually compiling
[05:53:18] <zwisk> right.
[05:54:37] <zwisk> And to be honest, I don't yet have a full grasp on the tcl bits to know what the significance of them sourceing 'bin/emcsh' or 'bin/iosh' and such directly is.... How do those get remapped?
[05:55:23] <zwisk> the paths are all relative, IIRC
[05:55:35] <jmkasunich> I don't know squat about tcl, sorry
[05:55:43] <zwisk> ok.. that's two of us. :)
[05:57:18] <alex_joni> make that 3;)
[05:57:33] <alex_joni> although I'm in the middle of compiling tcl on doze ;)
[05:57:49] <zwisk> I'm not sure why you do that to yourself, but... to each their own...
[05:58:05] <alex_joni> well..wanna see what I get out of emc on doze
[05:58:19] <alex_joni> jmk: did paul tell you I got TkEMC to run on doze?
[05:58:24] <alex_joni> connected to emc2?
[05:58:38] <jmkasunich> I don't remember if he did or not
[05:58:45] <alex_joni> well... I did ;)
[05:58:47] <alex_joni> last night
[05:58:53] <alex_joni> and it worked like a charm
[05:59:09] <alex_joni> with a doze zip file 2 years old
[05:59:18] <alex_joni> I was really stunned
[05:59:43] <zwisk> I'm stunned to. My luck with getting *anything* to work with windows is pitifully poor.
[06:01:05] <jmkasunich> I never did any doze programming... did DOS and 32-bit DOS with Watcom years ago, then quietly dropped out of programming until Linux came along
[06:01:36] <alex_joni> anyways...
[06:02:04] <alex_joni> I'm voting for zwisk to coordinate the make install thingy
[06:03:00] <jmkasunich> how about the make thingy? is that you? ;-)
[06:03:14] <jmkasunich> or should I say the configure thingy?
[06:03:26] <zwisk> I'm happy to take up the make install thing...
[06:03:38] <alex_joni> the configure thingy is ok for me ;)
[06:03:39] <zwisk> I'd like to graft on to that make rpm at some point...
[06:04:06] <zwisk> who's our resident tcl expert who can tell me what I've broken for tcl stuff?
[06:04:09] <alex_joni> well.. when make install is working ... we can go to make tgz, rpm, deb, srpm
[06:04:16] <jmkasunich> I like to see the basic premise be that "./configure ; make" builds a system that can be run and tested in place
[06:04:20] <jmkasunich> then make install installed it
[06:04:31] <zwisk> I like that a lot too...
[06:04:33] <jmkasunich> and later, make deb, make rpm, etc, can make packages
[06:04:34] <alex_joni> and make uninstall
[06:04:36] <alex_joni> :D
[06:04:43] <alex_joni> just in case there's an emc3
[06:04:51] <alex_joni> no other reason comes to mind...
[06:05:07] <zwisk> bah.. is there a make uninstall for windows? If they don't need it, we don't need it! :)
[06:05:52] <jmkasunich> make tgz is for a source tarball?
[06:06:09] <jmkasunich> if so, it is at least as important as make install, and more important than make rpm or deb
[06:06:47] <jmkasunich> after all, you can run a machine without make install, as long as run-in-place works
[06:06:53] <zwisk> make tgz is likely a pre-requisite for make rpm and make deb.
[06:07:18] <robin_sz> nah
[06:07:34] <robin_sz> rpms should be built from .spec files ..
[06:07:46] <zwisk> Yes, and spec files require SOURCE files...
[06:08:20] <robin_sz> yeah, but the instructions on what needs to go into the rpm and where it shoudl be exracted to, a lot of that is in the .spec
[06:08:35] <zwisk> Very true.
[06:09:02] <robin_sz> paul_c has some .spec files for tng I think ...
[06:09:25] <robin_sz> shake his tree, see if anything falls out.
[06:10:42] <zwisk> yep. I have spec files for emc2 as well.
[06:10:42] <zwisk> emc2 is cake compared to emc1.
[06:10:42] <picnet> * picnet wont mention turbocnc.
[06:13:26] <CIA-6> 03zwisk 07auto_configure_0_1 * 10emc2/Makefile: Fixes for installation problems relating to copying the 'CVS' directory
[06:13:35] <zwisk> CIA is still so cool!
[06:14:25] <jmkasunich> I was considering having the compile farm machines announce their results here too
[06:14:35] <zwisk> I think that's a really cool idea.
[06:14:57] <robin_sz> * robin_sz nods ...
[06:15:03] <jmkasunich> I'll look into it - after I fix the FTP problem
[06:15:14] <robin_sz> * robin_sz goes for a cup of tea
[06:15:19] <jmkasunich> (I'm planning on implementing timeouts and retries for FTP)
[06:16:04] <alex_joni> doesn't the client know how to do that?
[06:16:24] <jmkasunich> the client is command line ftp
[06:16:36] <jmkasunich> sometimes it just doesn't connect - could be comms problems, etc
[06:17:06] <jmkasunich> sometimes it exits with an error, other times it just waits (and waits, over 24 hours in one case, until I killed it)
[06:17:20] <jmkasunich> the existing script doesn't handle either case correctly
[06:17:24] <jmkasunich> but I'm working on it
[06:21:10] <alex_joni> I don't like dumb clients... who can't time-out
[06:24:59] <alex_joni> and I really hate windows compiling :((
[06:27:33] <alex_joni> error LNK2001: unresolved external symbol _mainCRTStartup
[06:27:40] <alex_joni> very usefull message ;)
[06:27:55] <alex_joni> ok.. let's resume more serious stuff:
[06:28:16] <alex_joni> zwisk: what can I help with make install?
[06:28:28] <alex_joni> what should I add to autoconf?
[06:28:52] <alex_joni> checks for awk, sed, ps, kill, whoami, etc?
[06:28:53] <zwisk> I'm at the moment adding to autoconf tests for as many of the programs as I can... So far, AC_PROG_INSTALL, and working on INSMOD and friends...
[06:28:55] <zwisk> yes...
[06:29:01] <zwisk> Would you like to take that on instead?
[06:29:18] <alex_joni> hmmm..thing is right now I can't really do nothing...
[06:29:32] <zwisk> Ok... I'll plod on then...
[06:29:34] <alex_joni> because I'm away from home, in some hotel...
[06:29:47] <zwisk> ahh.. fun :) Business, or pleasure? At least they have internet access!
[06:29:53] <alex_joni> but I'd like something left to do when I return
[06:30:03] <zwisk> ok. What shall I leave for you? :)
[06:30:04] <alex_joni> they have no fscking access
[06:30:11] <alex_joni> it's my CDMA cell phone ;)
[06:30:28] <zwisk> Hmm..that's pretty cool.
[06:30:30] <alex_joni> hmmm... dunno
[06:30:42] <alex_joni> what's not that urgent ...
[06:30:58] <alex_joni> I'll be back on thursday
[06:31:00] <zwisk> it's mostly cleanup. Paths are the 'interesting' part that gets hard for tcl, which we don't yet understand.
[06:31:02] <alex_joni> maybe a day sooner
[06:32:41] <alex_joni> hmmm.. I see tcl in $EMC_HOME/tcl
[06:33:01] <alex_joni> with /bin and /scripts as subdirs
[06:33:36] <zwisk> yep. The install map, however, doesn't currently include a 'scripts' directory... those all go in 'bin'. And there are relative references to 'scripts/' in various places.
[06:34:06] <alex_joni> don't thinks it's a good idea to move everything to bin
[06:34:20] <alex_joni> how bout moving tcl to 'bin'/scripts/
[06:35:05] <zwisk> erm... I don't really care, but when we discussed it in email on the dev list, the consensus was all scripts went to 'bin'. We can change that .. but ... ?
[06:35:09] <alex_joni> AFAIK tkemc.tcl is the interface
[06:35:23] <alex_joni> and it runs with emcsh
[06:35:47] <alex_joni> tcl's in bin/ are programs that can be run from the interface
[06:35:58] <alex_joni> need not to exist...I think tkemc can work without
[06:36:15] <jmkasunich> the current directory.map has scripts in their own directory. but I don't know if the install should do the same
[06:36:25] <jmkasunich> good questions for paul
[06:36:42] <paul_c> * paul_c returns after a power outage
[06:36:52] <jmkasunich> btw, don't hesitate to either email copies of the maps around for comment, or even to commit revsions for discussion
[06:37:00] <alex_joni> btw... jmk: remember I was complaining my tkemc started with small letters?
[06:37:07] <zwisk> the current one in there is defintely up for discussion :)
[06:37:10] <alex_joni> I talked to paul about that too
[06:37:10] <jmkasunich> vaguely
[06:37:31] <alex_joni> he said something about KDE which didn't fix the thing
[06:38:03] <alex_joni> I fixed it by copying TkEMC to the main dir
[06:38:39] <paul_c> net split...?
[06:39:35] <paul_c> alex_joni: Problems with fonts/colours and Tkemc ?
[06:40:15] <jmkasunich> well that's annoying
[06:40:22] <jmkasunich> KDE just up and died on me
[06:40:28] <jmkasunich> all I did was close a shell window
[06:40:54] <paul_c> Had that happen in the past
[06:41:32] <alex_joni> paul_c: I had them
[06:41:42] <alex_joni> fonts on tkemc where too small
[06:41:55] <alex_joni> when I went to settings-font it got fixed
[06:42:22] <alex_joni> I fixed the thing by copying TkTMC (notice the spelling) from emc1 to emc2
[06:43:43] <paul_c> TkEmc
[06:44:03] <paul_c> that is a config for fonts & colours
[06:44:50] <alex_joni> exactly
[06:45:31] <alex_joni> tkemc searches for the file at startup
[06:45:40] <alex_joni> foreach f {TkEmc configs/TkEmc /usr/X11R6/lib/X11/app-defaults/TkEmc} {
[07:10:36] <alex_joni> ok... guess I'm going to bed
[07:11:05] <alex_joni> maybe you guys can send a mail on the devel list with what has been done and what needs to be done
[07:11:15] <paul_c> catch you later Alex.
[07:11:20] <alex_joni> bye paul
[07:11:25] <jmkasunich> bye alex
[07:11:33] <jmkasunich> thanks for all your work
[07:11:42] <alex_joni> don't mention it
[07:23:39] <CIA-6> 03zwisk 07auto_configure_0_1 * 10emc2/ (configure configure.in configs/emc.conf.in scripts/emc.run):
[07:23:39] <CIA-6> emc.run now looks for emc.conf in a number of good-guess locations
[07:23:39] <CIA-6> configure now looks for grep, install, sed, ps, kill, whoami, awk,
[07:23:39] <CIA-6> insmod, rmmod, lsmod, piddof and ipcs at configure time and stores
[07:23:39] <CIA-6> those values in emc.conf rather than looking for them at runtime in
[07:23:40] <CIA-6> emc.run
[07:25:22] <paul_c> zwisk: A word in your lug if I may...
[07:25:40] <zwisk> lug?
[07:25:46] <jmkasunich> brit for ear
[07:26:01] <zwisk> ahh.. absolutely... I'm all ... uh... lugs...
[07:26:15] <paul_c> that's "all ears"
[07:26:23] <zwisk> :)
[07:26:59] <paul_c> as you add to or change stuff in the make files & configure.in
[07:27:16] <paul_c> can you add loads of comments please.
[07:27:34] <zwisk> sure. I'd be happy to.
[07:27:58] <jmkasunich> in the makefiles and .in files - you're already doing quite nicely at with your commit log messages ;-)
[07:27:59] <paul_c> Great - That will make life so much easier in the future
[07:28:31] <paul_c> commit logs are fine for "why a change was made"
[07:28:50] <jmkasunich> in-file comments explain how it works
[07:28:53] <jmkasunich> we need both
[07:28:53] <paul_c> notes in the code explain "wth this widget does"
[07:29:20] <zwisk> Can do.
[07:29:49] <paul_c> that will keep the BOFs happy
[07:30:05] <jmkasunich> BOFs?
[07:30:08] <zwisk> The makefile I certainly have not commented well enough yet. Are the configure.in comments not sufficient?
[07:30:15] <paul_c> Boring Old Farts
[07:31:12] <paul_c> configure.in comments should document the tests & rationale for them
[07:31:40] <paul_c> Makefile documents, well.. targets
[07:32:28] <paul_c> BTW did you like the {KVER} test ?
[07:33:15] <zwisk> yep, though I haven't stuck it in yet... For some odd reason, my kernel has '/usr/src/linux' in there, rather than /usr/src/linux-2.4.27' ... so... I gotta figure out what patched that up ...
[07:33:34] <paul_c> TNG ?
[07:33:42] <zwisk> nope. Redhat 9.
[07:34:18] <paul_c> /usr/src/linux is not a symlink to something else ?
[07:34:53] <zwisk> yes, but the wrong thing. It symlinks to my non-rtai kernel, even though I installed my rtai-kernel most recently. There's an ln and a sed scripts gone awry someplace.
[07:35:26] <paul_c> you been building kernel RPMs again ?
[07:35:55] <zwisk> yup. It's the only way to fly :)
[07:36:13] <zwisk> 2.4.27 and rtai-3.1 is much better than previous incarnations.
[07:36:35] <paul_c> with RTAI, the kernel source tree should be declared by the rtai-config script
[07:36:42] <zwisk> have you given up on rpms?
[07:36:54] <paul_c> * paul_c curses RPM dependencies
[07:36:58] <zwisk> Yes... and rtai-config also incorrectly reports /usr/src/rtai ...
[07:37:09] <zwisk> uhh.. sorry... /usr/src/linux
[07:37:19] <zwisk> what dependencies seem to get you?
[07:37:34] <paul_c> the fact there is no standards
[07:37:45] <zwisk> oh yes. That does cause pain.
[07:37:55] <zwisk> is dpkg better?
[07:38:08] <paul_c> any Tom Dick or Harry can set their own deps
[07:38:36] <zwisk> (and they do!)
[07:39:05] <paul_c> Debian is much better - but that is due to the fact that there is only one primary source
[07:39:28] <paul_c> third party debs can be just as bad as RPM
[07:39:58] <zwisk> I guess I usually consider redhat to be my 'one source'. And if I stray from that... well... I've brought the pain down on myself.
[07:39:59] <paul_c> * paul_c has a specific CADCAM package in mind....
[07:40:48] <paul_c> One problem with RPM - You can unpack a RH5.x on a Mandrake 10 install
[07:40:59] <paul_c> and break loads of things
[07:41:27] <zwisk> ooh.. I bet!
[07:41:47] <paul_c> Install a deb from Sid, and you can (up to a point) be assured it will work on a Woody box
[07:42:13] <zwisk> But, that's breaking the rules, IMHO. It's like untaring / from a solaris x86 box on your linux box. :)
[07:42:37] <zwisk> upgrading from rh5 via rpm to rh9 is similarly 'compatible'.
[07:42:55] <paul_c> You'd be suprised....
[07:43:29] <zwisk> I have a server that I've hat since redhat 6.x and it's now running redhat 9 after upgrading using rpms...
[07:43:32] <paul_c> Had someone install a RH7.x (as I recall) tcl package on a BDI-2.xx box
[07:44:05] <paul_c> and wondered why it wouldn't work
[07:44:59] <zwisk> yeah... well... People try to use computers without plugging them in and wonder why they don't work too.... but I can't blame the computer for *that*.
[07:47:31] <paul_c> don't you just lurv usrs
[07:48:02] <zwisk> Yeah... but it's human nature. They gotta learn somehow.
[07:52:16] <paul_c> zwisk: awk, grep, sed, & ps should all be in the usr's PATH
[07:52:46] <paul_c> and install, insmod, & kill in the root's PATH
[07:53:19] <zwisk> agreed... should be. But emc.run went through the trouble of looking for them, so I figure that configure should too.
[07:53:20] <paul_c> (do we need a check for ipcs ? )
[07:53:35] <zwisk> emc.run uses ipcs.
[07:53:41] <jmkasunich> configure should probably check that they are, and complain if not, but not actually store paths for them
[07:53:52] <zwisk> My reasoning is that I'd rather know that things can't be found in configure than at runtime.
[07:54:17] <paul_c> grep, sed, & awk are used during the configure/make
[07:54:30] <jmkasunich> emc(2).run copied what emc1.run did... I don't know if that's a good reason
[07:54:39] <zwisk> yes, they are.
[07:54:55] <paul_c> kill, ps, & whoami are (usually) part of a base install
[07:55:09] <zwisk> Well, if there's no 'good reason' for it, I'm happy to take all of that out of emc.run and let it just fail with a normal error if those executables aren't available.
[07:55:16] <jmkasunich> verify their availability in the normal user and root paths, yes, find them elsewhere and save the path, no
[07:55:48] <jmkasunich> I know for a fact that Ray would like to know at compile time if something he's gonna need a runtime is missing
[07:55:55] <zwisk> SO... that means undoing all of the variable call stuff that was used in emc.run ... which again, I'm fine with...
[07:56:08] <jmkasunich> declutters it a but
[07:56:10] <jmkasunich> bit
[07:56:17] <paul_c> which, bc, & dc are good candidates for checking for
[07:56:17] <zwisk> ok. Can do.
[07:57:33] <paul_c> would also suggest a check for sudo might be usefull
[07:58:51] <zwisk> Is whoami always in the path?
[08:00:12] <paul_c> should be
[08:24:09] <CIA-6> 03zwisk 07auto_configure_0_1 * 10emc2/Makefile: Cleanup and commenting of Makefile, per paul_c's request
[08:29:17] <robin_sz> * robin_sz returns
[08:32:23] <jmkasunich> * jmkasunich leaves
[08:32:29] <jmkasunich> (dinner time)
[08:32:36] <robin_sz> oh right
[08:32:44] <robin_sz> I thought it was my armpits :)
[08:47:12] <les> cough
[09:16:26] <asdfqwega> robin_sz, you still about?
[10:12:27] <paul_c> Looks like everything has died down...
[10:17:00] <jmkasunich> pretty much
[10:17:19] <jmkasunich> * jmkasunich is in the middle of repairing the dishwasher
[10:17:40] <paul_c> Coo...
[10:17:51] <paul_c> maybe I spoke too soon
[10:18:21] <Sendkeys> ell paul
[10:18:23] <Sendkeys> ello*
[10:21:55] <paul_c> So introduce yourself Sendkeys
[10:27:30] <Sendkeys> ahh was getting somthing to eat. well just new to emc trying to learn about it seeing if it will do what i need :)
[10:28:10] <Sendkeys> mostly trying to close the loop on my servo system without paying the big bucks
[10:28:38] <paul_c> A real servo or a G320 step/dir setup ?
[10:29:23] <Sendkeys> as of now im running rutex drives
[10:29:46] <paul_c> step/dir setup...
[10:29:50] <Sendkeys> yep
[10:30:23] <Sendkeys> was told pico-systems has a board to run emc then will turn it into a real system. basicly like a servotogo card without the 800 dollar price
[10:30:32] <paul_c> * paul_c finds it confusing when people talk about "servos" when using Rutex/Geckos
[10:30:33] <Sendkeys> about the only thing i have seen without replacing everything
[10:30:47] <Sendkeys> hehe
[10:31:18] <paul_c> Vital Systems do an interface card for a lot less than an STG
[10:31:32] <paul_c> and it is PCI
[10:31:50] <Sendkeys> ahhh cool
[10:32:30] <Sendkeys> 795 :P
[10:33:08] <Sendkeys> pico-system if it would even work.. still not 100% sure yet is only 250
[10:33:25] <paul_c> http://www.vitalsystem.com/motion/index.htm
[10:33:59] <Sendkeys> ahh 500 not bad
[10:34:02] <paul_c> <$500 for eight axis
[10:34:57] <Sendkeys> nice... but then again if i had 8 axis i wouldnt care about 500 bucks hehe
[10:35:39] <paul_c> Even for a three axis mill, a five axis card is usefull..
[10:35:51] <Sendkeys> true :)
[10:36:15] <paul_c> one for a rotary table, and the other for spindle
[10:37:14] <a70camaro> How close is EMC2 to being ready?
[10:37:50] <Sendkeys> good question :) would like to know that too haha
[10:38:25] <paul_c> * paul_c prods jmkasunich for the answer
[10:38:54] <a70camaro> other question , where to get a transformer to make a 120v DC power supply
[10:39:21] <paul_c> RS Components ?
[10:39:21] <Sendkeys> how many amps?
[10:40:12] <a70camaro> um , I have 4 servo 2amp continuous 10 amp stall 120v nominal volts
[10:40:37] <paul_c> The rating of the servo amps ?
[10:40:49] <a70camaro> overkill they are rated at 25 amp
[10:40:58] <paul_c> and voltage ?
[10:41:04] <a70camaro> they can go 180v
[10:41:11] <paul_c> Copley ?
[10:41:14] <a70camaro> AMC
[10:41:24] <paul_c> similar beast
[10:41:36] <a70camaro> i actually have apower supply that rectifies AC directly
[10:41:48] <a70camaro> no transformer , that scares me :-)
[10:42:10] <a70camaro> but it does 170v
[10:42:16] <paul_c> I was going to suggest an isolation transformer
[10:42:46] <Sendkeys> guy on ebay fill wind you one for like 150 bucks
[10:42:55] <Sendkeys> will*
[10:43:12] <Sendkeys> one sec let me get you the name
[10:43:32] <a70camaro> Ive been wondering if I could get by with just a isolation and keep the motor speed/voltage down
[10:43:54] <paul_c> A 4-6A transformer would probably be big enough for you..
[10:44:12] <a70camaro> i can limit encoder count/speed to keep it below 120 output , would this work ?
[10:45:44] <a70camaro> Ive got a spec on a power supply ,110VAC Primary / 83VAC & 24VAC Secondary Transformer 1.75KvA
[10:45:44] <a70camaro> 110VDC @ 25A after rectification
[10:46:02] <a70camaro> this would work I gues?
[10:46:29] <paul_c> 25A is way overkill for your needs
[10:46:57] <Sendkeys> johnango is his ebay name. if you want to get one comstum has nice prices
[10:47:05] <a70camaro> that's OK , ive been thinmking of using a dc servo for my lathe spindle
[10:47:29] <a70camaro> ive got 6 of these AMC amps ;-)
[10:48:08] <a70camaro> i will look him up and see , thanks sendkeys
[10:49:14] <Sendkeys> i ran into the problem as my shop didnt have 25amp 120volt wiring would make sure you can use that :) might have to go 220/240
[10:50:23] <Sendkeys> brb going to eat dinner
[10:50:35] <paul_c> hey guys - Nearly 1:45AM here
[10:50:47] <a70camaro> there is another power supply that is 220v input , it does 45 amps :-/ and its almost 500 bucks
[10:50:48] <paul_c> I'm going to have to leave you to it.
[10:51:03] <a70camaro> thanks
[10:51:07] <a70camaro> for the info
[10:59:22] <asdfqwega> paul_c: Still interested in a few etchings?
[10:59:50] <asdfqwega> ah, he's gone
[11:05:45] <jmkasunich> goodnight all (time to get some real work done)
[20:47:30] <paul_c> Morning sxpert